Covariate

A covariate is a continuous variable that is expected to change (“vary”) with (“co”) the outcome variable of a study. Generally speaking, a covariate can refer to any continuous variable that is [Page 283]expected to correlate with the outcome variable of interest. Although the term is sometimes used in this way, it is typically used to refer to variables that are not of direct or substantive interest in the study. Because they are associated with the outcome variable but are not the focus of the study, covariates are often called “extraneous” or “nuisance” variables. They are included in the analysis to increase precision or rule out alternative explanations for the findings, but they are not the focus of the analysis.

Accounting, or controlling, for covariates in analysis is important because it allows the researcher to be more confident in the conclusions drawn from the study and helps explain variance in the outcome variable that would otherwise be considered error variance. For example, in a study examining the effects of different group decision-making techniques on group members’ satisfaction with the group’s decision, it is likely that participants’ preexisting attitudes toward group work affect their satisfaction. The researcher could measure participants’ attitudes and include this variable as a covariate in his or her analyses. By doing so, the researcher could better isolate the effects of the different decision-making techniques and be more confident that any differences between groups were the result of those techniques, as opposed to preexisting attitudes about group work. The remainder of this entry describes typical uses of covariates and important considerations when including covariates in an analysis.